ClickHouse

Cisco Talos is one of the world’s largest and most trusted cybersecurity threat intelligence providers, running a hash reputation service that spans nearly 2 trillion rows and ~2 PB of data. Their old self-managed OSS ClickHouse setup worked, but for a small team, the operational burden, SOC 2 overhead, and architectural complexity were pulling engineers away from building. At Open House SF, Senay Goitom shared why the team chose ClickHouse Cloud, how they migrated with zero downtime, and what the new system has delivered: Read the full story 👇 https://lnkd.in/eT9n4bKb

ClickHouse open source turns 10 today! To the 2,000+ people who are building alongside us: thank you. A decade ago, Alexey Milovidov released a database built to handle web analytics at a scale no existing tool could manage. The goal was "Level 3" open source: transparent roadmap, open code review, public CI, full documentation, every contributor credited. He tells the whole story in the post below. From a scaling problem to a global project → * Started as a solution to 100 billion daily records across 500 columns * Open-sourced June 15, 2016, with a new brand and a public website * 2,000+ contributors and still growing * Every release changelog credits every contributor by name https://lnkd.in/eqkr_Vz2

Data is the ultimate building block for modern intelligence, but a foundation is only as strong as the infrastructure supporting it. I’m incredibly proud to see ClickHouse featured on the cover of CXO Insight Middle East this month! Our very own Arno van Driel sits down to discuss the critical foundations required to power true data intelligence and real-time analytics in today's rapidly evolving landscape. As organizations across the Middle East scale their digital transformation and AI initiatives, having a blazing-fast, reliable data architecture isn't just an advantage—it's a necessity. Check out the full May 2026 issue to read Arno's insights on how we are helping businesses build those robust foundations: https://lnkd.in/eiWDS3Di #DataAnalytics #RealTimeData #ClickHouse #MiddleEastTech #DataInfrastructure #CXOInsight José Luis Noriega Floortje Plokker

Earlier this year we redesigned the ClickHouse full-text index to run on object storage, where the random-access patterns that work on local disk are painfully slow on S3. We rebuilt it around sequential reads, with front-coded dictionary blocks, an in-memory sparse index, and posting lists that switch format based on how often a token appears. The key insight is that 94.5% of tokens appear in six rows or fewer, so we embed those posting lists directly in the dictionary. A 10M-row search dropped from 0.198s to 0.027s. https://lnkd.in/gjpn2jAF

Just a quick shout out to the ClickStack Demo Days. The team started recording these mostly for ourselves as a way to keep a record of what we're building each week, but we've ended up sharing them with the community too. They're not polished product launches, just a look at what we've been working on over the last few weeks. Apologies in advance for the occasional technical hiccup, interruption, or joke that only makes sense to the team. The last couple cover things like extended trace views, source-level filters, anomaly alerts, sharing investigations as notebooks, and a bunch of UX improvements around thresholds and color palettes. If you're curious about where ClickStack is heading, they're the up-to-date view of what the team is actually working on. 📺 May 29: https://lnkd.in/gCnC2-pK 📺 June 5: https://lnkd.in/g5X8v3Mn

We're excited to share that ClickHouse has achieved the Amazon Web Services (AWS) Retail Competency in the Advanced Data Insights category. Retail data is high-volume, high-cardinality, and high-concurrency, and the decisions it feeds expire fast: inventory, pricing, personalization, and retail media all run on what shoppers did seconds ago, with Black Friday peaks multiplying traffic exactly when slow dashboards cost the most. This recognition reflects customer success across retail and e-commerce, including: 🔹 Mercado Libre for full-fidelity marketplace observability and retail media analytics 🔹 Picnic Technologies for shift-by-shift visibility across 20+ automated fulfillment centers 🔹 Datavations and Rapid Delivery Analytics for CPG and retail market intelligence 🔹 Adevinta and Amp for seller analytics and merchant reporting Read the full announcement: https://lnkd.in/gMpBbVrG

We’re taking over San Francisco next week. ☕️⚡ The ClickHouse Cafe is open at the Contemporary Jewish Museum, right in front of Moscone, and we want to see you. Here’s what’s waiting for you: ☕ Espresso bar + housemade summer refreshers 🌮 Food trucks outside from 11am and again at happy hour 🧠 The ClickHouse Renovators Bar — our genius bar for your data stack questions 🟥 Langfuse Lounge for all things observability and AI tracing 🎮 A Lego set up for grabs! 🍻 Happy hour from 5–8pm We’re open 8am–8pm. Walk in, recharge, say hi, grab a drink, ask us anything. Register here: https://lnkd.in/g2-649Jr #DataAISummit #ClickHouse #Databricks

Did you hear about House Mates at Open House last month? It's the new ClickHouse partner community and program, launched with 60+ partners in the inaugural cohort. That includes 25+ ISVs with production-ready integrations, with dbt Labs, Fivetran, Confluent, Grafana Labs, Sigma, and Temporal among them, and 35+ consulting and services partners certified through ClickHouse Academy. The program runs three tiers based on the depth of the partnership: Ignite, Accelerate, and Prime. https://lnkd.in/gQa9XEbT

A year of ClickHouse benchmarks distilled into one post. The consistent thread across JSON performance, input formats, object storage caching, updates, and distributed aggregation: how efficiently does the engine move and process data end-to-end? A few things that stood out: - Native format wins ingestion across essentially every scenario tested. Netflix implemented native + LZ4 after these results; they now handle ~5PB/day. - Distributed object storage cache reduced cold query latency to 4× faster than local SSDs. - Standard SQL UPDATEs now run up to 1,000× faster than classic mutations — bulk operations that took minutes now take milliseconds. - Distributed GROUP BY processed 100 billion rows in ~0.4 seconds at ~240 billion rows/second. Performance depends on "how efficiently an engine moves and processes data end-to-end" — latency, batching, caching strategy, and execution model matter more than isolated optimizations. https://lnkd.in/gRFun5MP

We are so excited for our team to be having a great time helping hackers enjoy ClickHouse and Postgres 🔥